Recent Activity
Activity List
-
-
-
Welcome Aboard!April 22
-
-
-
-
-
-
-
-
-
-
Welcome Aboard!April 10
-
Welcome Aboard!April 9
-
Welcome Aboard!April 8
-
-
-
-
-
-
Im trying to carryout an ANOVA but im getting the error in the screenshot attached. I have also attached the excel sheet in which the dataset I am using is in.
Any help would be much appreciated
April 5 -
Im trying to carryout an ANOVA but im getting the error in the screenshot attached. I have also attached the excel sheet in which the dataset I am using is in.
Any help would be much appreciated
April 5 -
-
-
-
-
Part I
I think I will describe this situation as a "semi-bug", but unless one figures out the situation, it appears to be a bug.
I downloaded JASP 18.3 yesterday. The last time I used JASP was JASP14.1
It's important to say ahead of time that the dataset had no missing data. That's because I inserted an imputed value ahead of time at the initial stages of the dataset.
I uploaded an SEM program which I previously ran with 14.1. When I previously ran it with the dataset there were no problems. Tonight, however, I got the message
"FIML missing data handling only available with ML-Type estimators". Also, the ENTIRE OUTPUT WAS WIPED OUT.
Here are the choices I had chosen when I ran the data with 14.1
Estimators DWLS (I could not assume normality, which many researchers assume when they don't have normal distribution and have non-parametric data.)
Model Test Auto
Missing data handling FIML
Emulation None
Again, the model ran fine 2 years ago with 14.1, but tonight I got the error message
Solution: I changed FIML to Listwise deletion, and it ran fine.
Why did the solution work with 18.3? Answer : At first, the program said it can't use FIML for missing data unless I use maximum likelihood. So, I tried listwise deletion for missing data. Again, I knew I had NO missing data, so no subjects should be deleted. I don't know if other users are aware of this workaround. Perhaps, JASP can add the option: NO MISSING DATA. This would allow users a straightforward way of solving the problem.
Secondary problem, which is NOT a minor problem:
JASP automatically inserts the error message for ALL programs in the FILE. That is the JASP 14.1 output is wiped out for all the runs in the file. In order to retrieve the output for all the programs in the file, you have to rerun EACH program with the solution I suggested. This can take a long time, especially if you have an SEM model (confirmatory FA) with more than 100 items in a questionnaire with 11 hypothesized factors.
Is there another solution? Please Let me know.
PART II
For the syntax file window, please confirm
CTRL C means copy after highlighting
CTRL V means paste at the new cursor osition
IS R syntax identical to version 14.1 Lavaan syntax
I would appreciate a response relatively soon. I would like to go over my old output from 14.1 for other files, but I DON"T WANT TO WIPEOUT the 14.1 output when I want to read it with 18.3
Thank you.
Mark4
Another quick question:
Does JASP plan to include the Structure matrix from Principal Component Analysis together with the Pattern Matrix that it currently outputs with the oblique rotation option? JASP includes both Structure and Pattern matrices for the Exploratory FA, but not for Principal Components. SPSS produces both Structure and Pattern matrices for both Principal Components and Factor Analysis.There are reasons to have both ,matrices.
Thanks.
April 3 -
Part I
I think I will describe this situation as a "semi-bug", but unless one figures out the situation, it appears to be a bug.
I downloaded JASP 18.3 yesterday. The last time I used JASP was JASP14.1
It's important to say ahead of time that the dataset had no missing data. That's because I inserted an imputed value ahead of time at the initial stages of the dataset.
I uploaded an SEM program which I previously ran with 14.1. When I previously ran it with the dataset there were no problems. Tonight, however, I got the message
"FIML missing data handling only available with ML-Type estimators". Also, the ENTIRE OUTPUT WAS WIPED OUT.
Here are the choices I had chosen when I ran the data with 14.1
Estimators DWLS (I could not assume normality, which many researchers assume when they don't have normal distribution and have non-parametric data.)
Model Test Auto
Missing data handling FIML
Emulation None
Again, the model ran fine 2 years ago with 14.1, but tonight I got the error message
Solution: I changed FIML to Listwise deletion, and it ran fine.
Why did the solution work with 18.3? Answer : At first, the program said it can't use FIML for missing data unless I use maximum likelihood. So, I tried listwise deletion for missing data. Again, I knew I had NO missing data, so no subjects should be deleted. I don't know if other users are aware of this workaround. Perhaps, JASP can add the option: NO MISSING DATA. This would allow users a straightforward way of solving the problem.
Secondary problem, which is NOT a minor problem:
JASP automatically inserts the error message for ALL programs in the FILE. That is the JASP 14.1 output is wiped out for all the runs in the file. In order to retrieve the output for all the programs in the file, you have to rerun EACH program with the solution I suggested. This can take a long time, especially if you have an SEM model (confirmatory FA) with more than 100 items in a questionnaire with 11 hypothesized factors.
Is there another solution? Please Let me know.
PART II
For the syntax file window, please confirm
CTRL C means copy after highlighting
CTRL V means paste at the new cursor osition
IS R syntax identical to version 14.1 Lavaan syntax
I would appreciate a response relatively soon. I would like to go over my old output from 14.1 for other files, but I DON"T WANT TO WIPEOUT the 14.1 output when I want to read it with 18.3
Thank you.
Mark4
Another quick question:
Does JASP plan to include the Structure matrix from Principal Component Analysis together with the Pattern Matrix that it currently outputs with the oblique rotation option? JASP includes both Structure and Pattern matrices for the Exploratory FA, but not for Principal Components. SPSS produces both Structure and Pattern matrices for both Principal Components and Factor Analysis.There are reasons to have both ,matrices.
Thanks.
April 3 -
Welcome Aboard!March 29
-